1994
DOI: 10.1021/ac00087a012
|View full text |Cite
|
Sign up to set email alerts
|

Prediction of Reduced Ion Mobility Constants from Structural Information Using Multiple Linear Regression Analysis and Computational Neural Networks

Abstract: Multiple linear regression analysis and computational neural networks are used to develop models that predict reduced ion mobility constants (KO) from quantitative structural information encoded as descriptors. The errors associated with the models are similar to the calculated experimental error of -0,040 KO units. The best regression model contains five descriptors and has a multiple correlation coefficient (R) value of 0.991 and a standard deviation of 0.0469 KO units. The neural network model utilizes the … Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1

Citation Types

2
92
0

Year Published

1999
1999
2018
2018

Publication Types

Select...
7
2

Relationship

0
9

Authors

Journals

citations
Cited by 89 publications
(94 citation statements)
references
References 33 publications
2
92
0
Order By: Relevance
“…The CNNs used for this analysis are three-layered, fully connected, feed-forward networks, and they have been described in detail by Jurs and coworkers [41,42]. The number of neurons of the input layer corresponds to the number of descriptors in the model.…”
Section: Cnn Methods (Adapt)mentioning
confidence: 99%
See 1 more Smart Citation
“…The CNNs used for this analysis are three-layered, fully connected, feed-forward networks, and they have been described in detail by Jurs and coworkers [41,42]. The number of neurons of the input layer corresponds to the number of descriptors in the model.…”
Section: Cnn Methods (Adapt)mentioning
confidence: 99%
“…That is, CNNs that produce training and cross-validation set errors that are low and similar in magnitude tend to perform well in predicting the properties of interest for compounds not used in the training process. A quasi-Newton method BFGS (Broyden -Flectcher -Golfarb -Shanno) [42] is used to train the network. It should be noted that the ratio of training set observations to adjustable parameters should be kept above 2.0 to avoid overtraining [43].…”
Section: Cnn Methods (Adapt)mentioning
confidence: 99%
“…Whilst many approaches to IMS explicitly use or attempt to derive information on the 3D structure of 1217 the ion another approach is to use molecular descriptors to adequately describe an ion and predict 1218 the reduced mobility without any requirement to carry out computationally expensive geometry 1222 (Wessel and Jurs, 1994). Later, using six molecular descriptors on a training set of 135 compounds…”
mentioning
confidence: 99%
“…The CNNs used are three-layer, fully connected, feed-forward networks, that were employed in our previous papers on the pK a estimation of phenols [4] and benzoic acids [5], and have been described in detail by Jurs and coworkers [60,61]. The number of neurons of the input layer corresponds to the number of descriptors in the model.…”
Section: Cnn Methods (Adapt)mentioning
confidence: 99%